Internet Data Sheet
HYB18H256321BF
256-Mbit GDDR3
1.2
Description
The Qimonda 256-Mbit GDDR3 Graphics RAM is a high speed memory device, designed for high bandwidth intensive
applications like PC graphics systems. The chip’s 4 bank architecture is optimized for high speed.
HYB18H256321BF uses a double data rate interface and a 4n-pre fetch architecture. The GDDR3 interface transfers two 32
bit wide data words per clock cycle to/from the I/O pins. Corresponding to the 4n-pre fetch a single write or read access consists
of a 128 bit wide, one-clock-cycle data transfer at the internal memory core and four corresponding 32 bit wide, one-half-clock-
cycle data transfers at the I/O pins.
Single-ended unidirectional Read and Write Data strobes are transmitted simultaneously with Read and Write data respectively
in order to capture data properly at the receivers of both the Graphics SDRAM and the controller. Data strobes are organized
per byte of the 32 bit wide interface. For read commands the RDQS are edge-aligned with data, and the WDQS are center-
aligned with data for write commands.
The HYB18H256321BF operates from a differential clock (CLK and CLK). Commands (addresses and control signals) are
registered at every positive edge of CLK. Input data is registered on both edges of WDQS, and output data is referenced to
both edges of RDQS.
In this document references to “the positive edge of CLK” imply the crossing of the positive edge of CLK and the negative edge
of CLK. Similarly, the “negative edge of CLK” refers to the crossing of the negative edge of CLK and the positive edge of CLK.
References to RDQS are to be interpreted as any or all RDQS<3:0>. WDQS, DM and DQ should be interpreted in a similar
fashion.
Read and write accesses to the HYB18H256321BF are burst oriented. The burst length is fixed to 4 and 8 and the two least
significant bits of the burst address are “Don’t Care” and internally set to LOW. Accesses begin with the registration of an
ACTIVATE command, which is then followed by a READ or WRITE command. The address bits registered coincident with the
ACTIVATE command are used to select the bank and the row to be accessed. The address bits registered coincident with the
READ or WRITE command are used to select the bank and the column location for the burst access. Each of the 4 banks
consists of 4096 row locations and 512 column locations. An AUTO PRECHARGE function can be combined with READ and
WRITE to provide a self-timed row precharge that is initiated at the end of the burst access. The pipe lined, multibank
architecture of the HYB18H256321BF allows for concurrent operation, thereby providing high effective bandwidth by hiding row
precharge and activation time.
The “On Die Termination” interface (ODT) is optimized for high frequency digital data transfers and is internally controlled. The
termination resistor value can be set using an external ZQ resistor or disabled through the Extended Mode Register.
The output driver impedance can be set using the Extended Mode Register. It can either be set to ZQ / 6 (auto calibration) or
to 35, 40 or 45 Ohms.
Auto Refresh and Power Down with Self Refresh operations are supported.
An industrial standard PG–TFBGA–136 package is used which enables ultra high speed data transfer rates and a simple
upgrade path from former DDR Graphics SDRAM products.
Rev. 0.80, 2007-09
4
09132007-07EM-7OYI
QIMONDA [ QIMONDA AG ]
